feat(SQO): Introduce circuit breaker.

.gitignore

@@ -20,3 +20,6 @@ contracts/contract.abi.json
 /src/utils/__pycache__
 /src/models/__pycache__
 .coverage
+
+# Ignore docker compose override file
+docker-compose.override.yml

Dockerfile

@@ -1,7 +1,7 @@
 # Dockerfile to create a clean, lightweight Docker Image for the Service Quality Oracle
 
 # Use Python 3.9 slim as the base image for a lightweight container
-FROM python:3.9-slim
+FROM python:3.11-slim
 
 # Add metadata labels
 LABEL description="Service Quality Oracle" \

README.md

@@ -49,17 +49,21 @@ Please refer to the [ELIGIBILITY_CRITERIA.md](./ELIGIBILITY_CRITERIA.md) file to
 
 The application follows a clear data flow, managed by a daily scheduler:
 
-1.  **Scheduler (`scheduler.py`)**: This is the main entry point. It runs on a schedule (e.g., daily), manages the application lifecycle, and triggers the oracle run. It is also responsible for catching up on any missed runs.
+1. **Scheduler (`scheduler.py`)**: This is the main entry point. It runs on a schedule (e.g., daily), manages the application lifecycle, and triggers the oracle run. It is also responsible for catching up on any missed runs.
 
-2.  **Orchestrator (`service_quality_oracle.py`)**: For each run, this module orchestrates the end-to-end process by coordinating the other components.
+2. **Orchestrator (`service_quality_oracle.py`)**: For each run, this module orchestrates the end-to-end process by coordinating the other components.
 
-3.  **Data Fetching (`bigquery_provider.py`)**: The orchestrator calls this provider to execute a configurable SQL query against Google BigQuery, fetching the raw indexer performance data.
+3. **Data Fetching (`bigquery_provider.py`)**: The orchestrator calls this provider to execute a configurable SQL query against Google BigQuery, fetching the raw indexer performance data.
 
-4.  **Data Processing (`eligibility_pipeline.py`)**: The raw data is passed to this module, which processes it, filters for eligible and ineligible indexers, and generates CSV artifacts for auditing and record-keeping.
+4. **Data Processing (`eligibility_pipeline.py`)**: The raw data is passed to this module, which processes it, filters for eligible and ineligible indexers, and generates CSV artifacts for auditing and record-keeping.
 
-5.  **Blockchain Submission (`blockchain_client.py`)**: The orchestrator takes the final list of eligible indexers and passes it to this client, which handles the complexities of batching, signing, and sending the transaction to the blockchain via RPC providers with built-in failover.
+5. **Blockchain Submission (`blockchain_client.py`)**: The orchestrator takes the final list of eligible indexers and passes it to this client, which handles the complexities of batching, signing, and sending the transaction to the blockchain via RPC providers with built-in failover.
 
-6.  **Notifications (`slack_notifier.py`)**: Throughout the process, status updates (success, failure, warnings) are sent to Slack.
+6. **Notifications (`slack_notifier.py`)**: Throughout the process, status updates (success, failure, warnings) are sent to Slack.
+
+## Architecture
+
+For a more detailed explanation of key architectural decisions, such as the RPC provider failover and circuit breaker logic, please see the [Technical Design Document](./docs/technical-design.md).
 
 ## CI/CD Pipeline
 
@@ -128,7 +132,6 @@ bandit -r src/
 ## TODO List (only outstanding TODOs)
 
 ### 1. Testing
-- [ ] Create integration tests for the entire pipeline
 - [ ] Security review of code and dependencies
 
 ### 2. Documentation

docker-compose.yml

@@ -51,7 +51,7 @@ services:
           memory: 512M
 
     # Restart policy
-    restart: unless-stopped
+    restart: "on-failure"
 
     # Healthcheck to ensure the container is running
     healthcheck:

docs/technical-design.md

@@ -0,0 +1,56 @@
+# Technical Design & Architecture
+
+This document outlines key architectural decisions and data flows within the Service Quality Oracle.
+
+## RPC Provider Failover and Circuit Breaker Logic
+
+The application is designed to be resilient to transient network issues and RPC provider failures. It uses a multi-layered approach involving internal retries, provider rotation, and an application-level circuit breaker to prevent catastrophic failures and infinite restart loops.
+
+The following diagram illustrates the sequence of events when all RPC providers fail, leading to a single recorded failure by the circuit breaker.
+
+```mermaid
+sequenceDiagram
+    # Setup column titles
+    participant main_oracle as service_quality_oracle.py
+    participant blockchain_client as blockchain_client.py
+    participant circuit_breaker as circuit_breaker.py
+    participant slack_notifier as slack_notifier.py
+
+    # Attempt function call
+    main_oracle->>blockchain_client: batch_allow_indexers_issuance_eligibility()
+
+    # Describe failure loop inside the blockchain_client module
+    activate blockchain_client
+    alt RPC Loop (for each provider)
+
+        # Attempt RPC call 
+        blockchain_client->>blockchain_client: _execute_rpc_call() with provider A
+        note right of blockchain_client: Fails after 5 retries
+
+        # Log failure
+        blockchain_client-->>blockchain_client: raises ConnectionError
+        note right of blockchain_client: Catches error, logs rotation
+
+        # Retry RPC call
+        blockchain_client->>blockchain_client: _execute_rpc_call() with provider B
+        note right of blockchain_client: Fails after 5 retries
+
+        # Log final failure
+        blockchain_client-->>blockchain_client: raises ConnectionError
+        note right of blockchain_client: All providers tried and failed
+    end
+
+    # Raise error back to main_oracle oracle and exit blockchain_client module
+    blockchain_client-->>main_oracle: raises Final ConnectionError
+    deactivate blockchain_client
+
+    # Take note of the failure in the circuit breaker, which can break the restart loop if triggered enough times in a short duration
+    main_oracle->>circuit_breaker: record_failure()
+
+    # Notify of the RPC failure in slack
+    main_oracle->>slack_notifier: send_failure_notification()
+
+    # Document restart process
+    note right of main_oracle: sys.exit(1)
+    note right of main_oracle: Docker will restart. CircuitBreaker can halt via sys.exit(0) 
+```

requirements.txt

@@ -2,43 +2,42 @@
 # Requires Python 3.9+
 
 # Configuration management
-tomli==2.2.1  # TOML support for Python < 3.11
+tomli==2.2.1
 
 # Scheduling and resilience
 schedule==1.2.2
 pytz==2025.2
 tenacity==8.5.0
 
 # Google Cloud BigQuery for data processing
-google-cloud-bigquery==3.26.0
-bigframes==1.42.0
+google-cloud-bigquery==3.34.0
+bigframes==2.8.0
 
 # Data processing and validation
 pandas==2.2.3
 pandera==0.20.4
-numpy>=2.0.0  # Added as pandas dependency
+numpy==2.3.1
 
 # Blockchain integration - Latest compatible versions
 web3==7.12.0
-eth-account>=0.13.0
-eth-typing>=5.2.0
+eth-account==0.13.7
+eth-typing==5.2.1
 
 # GraphQL and subgraph integration (for future subgraph functionality)
 gql==3.5.2
 
 # HTTP and API
 requests==2.32.3
-aiohttp>=3.9.0  # For async HTTP requests (used by web3)
+aiohttp==3.12.13
 
 # Development/Testing
-pytest>=8.0.0
-pytest-cov>=6.0.0
-pytest-mock>=3.0.0
-pytest-snapshot>=0.9.0
-mypy>=1.0.0
-types-pytz  # Type stubs for pytz
-types-requests  # Type stubs for requests
+pytest==8.4.1
+pytest-cov==6.2.1
+pytest-mock==3.14.1
+pytest-snapshot==0.9.0
+mypy==1.16.1
+types-pytz==2025.2.0.20250516
+types-requests==2.32.4.20250611
 
 # Linting and formatting
-ruff>=0.6.0
-pip==25.1
+ruff==0.12.0